com.atlassian.confluence.content.render.xhtml.migration.exceptions.UnknownMacroMigrationException: The macro 'next_previous_link3' is unknown.

Deploying Streaming Applications

A Siddhi application (.siddhi file) can be deployed to run that on production. 

There are two modes deployment 

  • Standard deployment 
    Here we deploy Siddhi Apps in worker node (In Single node, Minimum HA Deployments)
  • Fully distributed deployment
    Here we deploy Siddhi Apps in manager node (Only in Fully Distributed Deployment) 

Refer below sections to get to know how to deploy on the above deployments:

Standard Deployment  

To start a worker node in single node mode, issue one of the following commands:

  • For Windows:   worker.bat
  • For Linux :   ./worker.sh

Siddhi applications can be deployed to the worker node by using one of the following methods:

  • Using Stream Processor Studio
    This involves deploying the Siddhi applications via the Stream Processor studio once you create and save them. To do this, follow the substeps below.

    This method allows you to deploy multiple Siddhi applications to mutiple servers at once.


    1. Open the Stream Processor Studio. For detailed instructions, see Stream Processor Studio Overview.
    2. In the top menu bar, click Deploy and then click Deploy to Server.
    3. If you want to deploy all the Siddhi applications saved in the Stream Processor Studio, select the check box for the Workspace directory as shown in the example below. If not, select the check boxes of the required Siddhi applications.
    4. Add the servers to which you want to deploy the Siddhi applications as follows:

      1. In the Host field, enter the host of the server.
      2. In the Port field, enter the port of the server.
      3. In the User Name field enter the user name to log in to the server.
      4. In the Password field, enter the password to  log in to the server.
      5. Click Add to add the server.

      Repeat the above substeps to specify all the servers to which the Siddhi applications you selected need to be deployed.

    5. Once you have added all the required servers, click Deploy. As a result, a log appears to indicate whether the selected Siddhi applications are successfully deployed as shown in the example below.
  • Deploying manually
    This involves dropping the .siddhi file in to the <SP_HOME>/wso2/worker/deployment/siddhi-files/ directory before or after starting the worker node.
  • Via REST API
    This involves sending a "POST" request to http://<host>:<port>/siddhi-apps with the Siddhi App included in the body of the request. Refer Stream Processor REST API Guide for more information on using WSO2 Strean Processor APIs.

When a Siddhi application is successfully deployed, a message similar to the following example appears in the startup logs.

To configure a Minimum HA deployment refer Minimum High Availability Deployment documentation.

Fully Distributed Deployment

To successfully set up, configure and run Siddhi applications in a fully distributed environment refer Fully Distributed Deployment documentation.

If you need to deploy streaming applications in a Docker environment, they need to be exported as Docker artifacts. For more information, see Exporting Siddhi Files as Docker Artifacts.




com.atlassian.confluence.content.render.xhtml.migration.exceptions.UnknownMacroMigrationException: The macro 'next_previous_links2' is unknown.